Interview Questions › Metrics

Are CRED coins rewards worth the cost?

CRED rewards users with coins when they pay credit card bills. Rewards cost a lot. How would you measure whether rewards are worth it, and what would you change?

Context

CRED earns from commerce, lending and merchant partnerships. Bill payment is the habit that brings users back each month.

Clarifying questions to ask first

• What share of revenue comes from users who engage with rewards?
• Can we run a reduced-rewards holdout?

How to approach it

1. What rewards are meant to cause: monthly bill payment habit
2. Incrementality: would users pay via CRED anyway?
3. Downstream revenue per rewarded user
4. Cost per incremental retained user

Framework: Goal of the product -> Users and value exchange -> North Star -> Supporting metrics by funnel stage -> Guardrails and counter-metrics -> Decision or target

Sample answer

• Incremental monthly bill payers vs reduced-reward holdout
• Reward cost per incremental payer
• Commerce and lending revenue per rewarded user

1. A holdout with half the rewards kept 93% of bill payers
2. Users who redeem coins in the store generate 4x revenue of those who do not
3. Reward cost per user is 3x higher for users who never redeem

Most bill payment is not caused by rewards, so cut blanket rewards and redirect spend to users likely to redeem in commerce. Measure cost per incremental monthly payer and revenue per rewarded user, with bill payer retention as the guardrail.

What interviewers look for

• Monthly bill payers
• Reward cost per incremental payer
• Revenue per user